Kaonic atoms and in-medium K-N amplitudes II: interplay between theory and phenomenology

0.24 MB·English

by  E. Friedman

Kaonic atoms and in-medium K N amplitudes II:
−
Interplay between theory and phenomenology
E. Friedmana, A. Gala
3 aRacah Institute of Physics, The Hebrew University, Jerusalem 91904, Israel
1
0
2

n
a
J Abstract

3
(1)
A microscopic kaonic-atom optical potential V is constructed, using
2 K−
theIkeda-Hyodo-Weise NLO chiral K−N subthreshold scattering amplitudes

]
h constrained by the kaonic hydrogen SIDDHARTA measurement, and incor-
-t porating Pauli correlations within the Waas-Rho-Weise generalization of the
l
c Ericson-Ericson multiple-scattering approach. Good fits to kaonic atom data
u −
over the entire periodic table require additionally sizable K NN–motivated
n
[ absorptiveanddispersive phenomenologicalterms, inagreementwithourfor-

3 mer analysis based ona post-SIDDHARTAin-medium chirally-inspired NLO
v separable model by Cieply´ and Smejkal. Such terms are included by intro-
6
(2)
ducing a phenomenological potential V and coupling it self-consistently to
3 K−
3 (1)
V . Properties of resulting kaonic atom potentials are discussed with spe-
6 K−
−
. cial attention paid to the role of K -nuclear absorption and to the extraction
1
−
1 of density-dependent amplitudes representing K multi-nucleon processes.
2
1 Keywords: meson-nuclear multiple scattering, meson-baryon coupled
: channel chiral models, kaonic atoms
v
i PACS: 13.75.Jz, 21.65.Jk, 36.10.Gv
X
r
a
1. Introduction
Thewealthofstrong-interactiondatadeducedfromkaonicatomsprovides
−
invaluable information on the K -nuclear interaction at threshold [1, 2]. Re-
−
cent studies of K atoms focused on constructing self-consistently a density-
−
dependent K -nuclear potential:
2ωKVK(1−)(ρ) = −4πF˜K−N(pρ,√sρ)ρ . (1)
Preprint submitted to Nuclear Physics A January 24, 2013
−
Here ω m at threshold, p~ is the in-medium relative K N momen-
K K ρ
≈
tum and s = (E + E )2 (p~ + p~ )2 is the in-medium Lorentz in-
ρ K N K N
−
variant Mandelstam energy-squared variable, both of which depend on the
density ρ [3, 4, 5]. The in-medium scattering amplitude ˜K−N(pρ,√sρ) re-
F
˜
duces in the low-density limit to fK−N(p=0,√s=mK+mN), the free-space
− −
K N scattering length in the K nucleus c.m. system. The dependence of
˜K−N(pρ,√sρ) on its kinematical variables was transformed in these stud-
F
ies, using a self-consistent procedure, into a density-dependent ˜K−N(ρ).
F
(1)
The resulting density-dependent optical potential V (ρ) accounts forsingle-
K−
nucleon K¯N K¯N elastic and K¯N πY reaction processes. Empirically,
¯ → →
KN amplitudes at about 40 MeV below threshold are involved in pinning
(1)
down V in kaonic atoms at half nuclear-matter density [3, 4, 5].
K−
(1)
This microscopic construction of V , however, did not provide any rea-
K−
sonable reproduction of the experimental values of strong-interaction level
shifts and widths in kaonic atoms. In particular, because of the rapid de-
˜
crease of the underlying absorptivity Im fK−N(p,√s) when the free-space
˜ − (1)
amplitudes fK−N are evaluated further below the K N threshold, Im VK−
was unable to account for the strong absorptivity content of kaonic atoms
(i.e. their level widths). Thus, in addition to the single-nucleon terms on the
r.h.s. of Eq. (1), a sizable phenomenological absorptive term together with a
strong dispersive term appeared necessary in order to achieve reasonable fits
to the data. This has been demonstrated recently in Refs. [3, 4, 5] where in-
− ˜
medium K N amplitudes K−N constructed within the chirally motivated
F
separable-interaction models of Cieply´ and Smejkal (CS) [6, 7] were used
(1)
to evaluate V . In particular, models TW1 and NLO30 from Ref. [7], ac-
K−
counting for the recently measured SIDDHARTA values of kaonic hydrogen
1s shift and width [8, 9], have been used in Refs. [4] and [5] respectively. It
(2)
was found in these works that the added V part consisting of dispersive
K−
(1)
and absorptive terms was as important as V .
K−
The present work is a natural extension of our recent work [5] which
may be considered part I of ongoing studies of kaonic atoms and in-medium
−
K N scattering amplitudes. The emphasis of the present work is on the
interplay between theory and phenomenology that emerges in kaonic atom
− −
studies. Starting from the free-space NLO chiral K p and K n s-wave
amplitudes constructed by Ikeda, Hyodo and Weise (IHW) [10, 11] which
account for the SIDDHARTA data, we arrive at similar conclusions to those
2
reached in part I outlined above. We note that the IHW construction is free
of any phenomenologically adjusted momentum-space form factors which in
the CS separable-model construction are not directly guided by a systematic
−
chiral hierarchy. The IHW free-space charge-averaged K N c.m. scattering
−
amplitude fK−N(√s) is shown in Fig. 1 below the K N threshold. One
notes its strong energy dependence, with RefK−N(√s) mostly rising in going
subthreshold and Im fK−N(√s) sharply dropping below √s = 1415 MeV as
one approaches the πΣ threshold (√s = 1330 MeV).
1.2
IHW average amplitude
Imag.
1.0
m) 0.8 Real
(f
e f 0.6
g
a
r
e 0.4
v
a
0.2
0.0
−0.2
1340 1360 1380 1400 1420 1440
s1/2 (MeV)
Figure 1: The K−N c.m. scattering amplitude fK−N(√s) = 12(fK−p(√s)+fK−n(√s))
− −
below threshold, constructed from the IHW free-space K p and K n s-wave scattering
amplitudes [11]. To convert from the two-body c.m. to the lab system, which for A 1
coincides with the K− nucleus c.m. system, apply f˜=(√s/m )f. ≫
N
˜
To generate in-medium amplitudes K−N(√sρ) from the IHW free-space
− − F
K p and K n s-wave amplitudes, we apply the multiple-scattering (MS)
approach of Waas, Rho and Weise (WRW) [12] focusing on Pauli correla-
tion effects, as described in Section 2. Charge (isospin) degrees of freedom
are incorporated in this MS approach which determines, under a straightfor-
ward generalization of Eq. (1), the single-nucleon density-dependent poten-
(1)
tial V (ρ). To represent multi-nucleon dispersive and absorptive processes,
K−
(2)
we add a phenomenological density-dependent interaction potential V (ρ).
K−
(1) (2)
Both V and V are coupled implicitly within a self-consistent cycle built
K− K−
3
into the kaonic atom fitting procedure. Properties of the resulting kaonic
atom potentials are discussed in Section 3 with special emphasis placed on
−
the role of K nuclear absorption. The paper ends with a brief summary in
Section 4.
2. Methodology
Several subjects are introduced and discussed in this section. In Sec-
tion 2.1 we briefly review the multiple-scattering procedure applied by WRW
[12] to incorporate nuclear-correlation contributions, particularly Pauli cor-
− (1)
relations in the construction of a K nucleus potential V at low energies
K−
in terms of in-medium K−N amplitudes ˜K−N(√sρ). In Section 2.2 we
F
(2)
introduce a phenomenological potential V to represent multi-nucleon pro-
K−
(1)
cesses outside of the underlying meson-baryon chiral model on which V is
K−
(1) (2)
based. Although no explicit coupling between V and V is practised in
K− K−
the present calculations, in agreement with the spirit of the original Ericson-
Ericson MS procedure [14], possible alternatives are considered in Appendix
A to this work. In Section 2.3 we focus attention on the self-consistent proce-
dure of relating subthreshold energies to densities through functional depen-
dence involving both V(1) and V(2), thereby coupling implicitly V(1) to V(2).
K− K− K− K−
Finally, in Section 2.4 we discuss the effect of the Λ(1405) subthreshold res-
onance on the low-density behavior of our in-medium amplitudes ˜K−N(ρ).
F
2.1. Overview of WRW
In this MS procedure one starts by relating the meson wavefunction φ(~r)
~
generated by the incident plane wave exp(ik ~r) to the effective meson wave-
·
functions φeff(r~′) at the scattering point r~′:
I
exp(ik ~r r~′ )
φ(~r) = exp(i~k ~r)+ d3r′ | − | f˜ρ (r~′)φeff(r~′) . (2)
· ~r r~′ I I I
IX=0,1Z | − |
Here, f˜ are free-space K¯N s-wave scattering amplitudes with good isospin
I
in the lab system and ρ = (2I+1)ρ/4 for isospin-symmetric nuclear density
I
ρ. Operatingwith(∆+k2)onbothsides of(2)oneobtainsthewave equation
(∆+k2) φ(~r) = 4π f˜ρ (~r)φeff(~r) . (3)
− I I I
I=0,1
X
4
The MS procedure relates effective wavefunctions φeff(~r) at ~r to effective
I
wavefunctions φeff(r~′) at r~′ as follows:
I′
exp(ik ~r r~′ )
φeIff(~r) = φ(~r)+ d3r′ ~r | r~−′ | cII′(~r,r~′)f˜I′ρI′(r~′)φeIff′ (r~′) , (4)
IX′=0,1Z | − |
where cII′(~r,r~′) are isospin projected NN correlation functions which in nu-
clear matter depend on t = ~r r~′ only. Replacing in the long-wavelength
| − |
limit the argument r~′ by ~r in ρI′(r~′) and φeIff′ (r~′), Eq. (4) reduces to
φeIff(~r) = φ(~r)− ξII′f˜I′ρI′(~r)φeIff′ (~r) , (5)
I′=0,1
X
where
∞
ξII′ = 4π dtexp(ikt)tcII′(t) . (6)
−
Z0
The following discussion is limited to Pauli correlation contributions to cII′
which were found by WRW (and also confirmed by us) to overshadow con-
tributions from dynamical short-range correlations. For isospin symmetric
matter, the Pauli ξII′ are diagonal in isospin, see Ref. [12]. Solving Eqs. (5)
for φeff(~r) in terms of φ(~r) and substituting on the r.h.s. of Eq. (3), the latter
I
(1)
assumes the form of 2ω V (ω ;ρ) φ(~r), with
K K− K
˜
2I +1 f
(1) I
2ω V (ω ;ρ) = 4π ρ(r) , (7)
K K− K − 4 1+ 1ξ f˜ρ(r)
I 4 k I
X
where ξ is given by
k
∞
9π dt
ξ = 4 exp(iqt)j2(t) , q = k/p . (8)
k p2 t 1 F
F (cid:18) Z0 (cid:19)
For kaonic atoms, k = (ω2 m2 )1/2 0 and ξ = 9π/p2, where the Fermi
K− K ≈ k=0 F
momentum p isgiven by p = (3π2ρ/2)1/3. We notethat thedensity depen-
F F
(1)
dence of V (ω ;ρ) in Eq. (7) is not limited to the explicit dependence on ρ
K− K
in its right-hand side, but arises as well from the √s energy argument of the
˜
free-space amplitudes f which for simplicity was disregarded in the deriva-
I
tionabove andwhich gives riseinthenuclear medium toadensity-dependent
subthreshold energy argument √s . The precise definition of √s and the
ρ ρ
5
˜
self-consistent procedure by which the √s dependence of f isconverted into
ρ I
a density dependence of F˜K−N and of VK(1−) are relegated to Section 2.3.
Eq. (7) was derived assuming implicitly equal proton and neutron density
distributions, ρ (r)=ρ (r)=ρ(r)/2. Relaxing this constraint, primarily for
p n
use in kaonic atom applications where ω µ with µ the reduced mass of
K K K
≈
the kaon, one obtains to leading order:
2µ V(1)(ρ) = 4π (2f˜K−p −f˜K−n) 21ρp + f˜K−n(21ρp +ρn) . (9)
K K− − " 1+ 41ξk=0f˜0ρ(r) 1+ 41ξk=0f˜1ρ(r)#
This form of the MS summation is used in all of the numerical applications
reported in the present work. Eq. (9) may be rewritten as
A 1 µ
(1) K
2µKVK−(ρ) = −4π (1+ A− m )[FK−p(ρ)ρp(r)+FK−n(ρ)ρn(r)] , (10)
N
which defines the in-medium amplitudes K−p(ρ) and K−n(ρ). Here, A is
F F
the atomic mass number. Finally, to bring Eq. (10) into the more compact
form (1), we define an effective amplitude eff (ρ) through
FK−N
FKeff−N(ρ)ρ(r) = FK−p(ρ)ρp(r)+FK−n(ρ)ρn(r) , (11)
−
with the standard conversion from two-body c.m. to the K -nuclear c.m.
frame given by ˜eff (ρ) = (1 + A−1 µK ) eff (ρ). Eq. (11) may serve for
FK−N A mN FK−N
defining similarly an effective free-space subthreshold amplitude feff (ρ),
K−N
with density dependence arising from the underlying energy dependence
√s √s ρ.
ρ
→ →
2.2. Adding V(2)(ω ;ρ)
K− K
We wish now to account for multi-nucleon dispersive and absorptive pro-
(2)
cesses by adding a phenomenological term V (ω ;ρ) to the MS single-
K− K
(1)
nucleon potential V (ω ;ρ), Eq. (7). Traditionally, these processes are not
K− K
iterated within the MS expansion in which scattering occurs on single nu-
cleons via amplitudes f˜, and V(2) is simply added to the MS V(1) specified
K− K−
in Eq. (7) by ξ = 9π/p2 for kaonic atoms. Alternative ways of intro-
k=0 F
(2)
ducing V , by letting it affect explicitly the MS procedure for deriving
K−
(1) (1)
the in-medium VK−, or by redefining the splitting of VK− into its VK− and
(2)
V components, are discussed briefly in Appendix A to this work. None of
K−
these alternatives was found in the present work to offer advantage over the
(2)
straightforward addition of V .
K−
6
2.3. Energy vs. density
Hereweoutlineabasicdifferencebetween thepresent workandourprevi-
−
ous ones [3, 4, 5] in which a model of using K N amplitudes below threshold
−
was introduced. In these calculations, energies and momenta of the K me-
son and a bound nucleon were determined, independently, by the nuclear
environment, and the Mandelstam energy variable √s was evaluated in the
nuclear medium, thereby becoming density dependent, √s . In particular,
ρ
−
the in-medium K momentum was assigned locally to the real part of the
− (1)
single-nucleon K -nucleus potential Re V , leading to
K−
√s E B ξ B 15.1(ρ/ρ )2/3 +ξ (Re V(1) +V ) (MeV) (12)
ρ ≈ th − N − N K − 0 K K− c
with E = m +m , ξ = m /(m +m ), ξ = m /(m +m ), and
th N K N N N K K K N K
−
with V for the K Coulomb potential, B for a nucleon average binding
c N
energy and ρ for nuclear matter density. The atomic binding energy B
0 K
−
of the K is relatively small and, hence, it is safe to neglect it in Eq. (12).
With a fixed value for B it is evident that for ρ 0 the energy approaches
N
→
E B , thus violating the low-density limit whereby the 1N amplitude
th N
−
should approach the free amplitude at threshold, E . In the present work
th
we have therefore replaced the fixed average nucleon binding energy B by
N
a density-dependent one,
B (ρ) = B ρ/ρ¯, (13)
N N
where the average density is given by
1
ρ¯= ρ2 d3r (14)
A
Z
with A the atomic mass number. For B we used the value of 8.5 MeV, the
N
same as in our earlier work. Furthermore, in order to enhance the very slow
convergence of √s to E due to the Coulomb potential, V was multiplied
ρ th c
in Eq. (12) by (ρ/ρ )1/3, based on dimensional arguments, thereby ensuring
0
that the low-density limit is respected in the calculations.
Since VK(1−) is proportional locally according to Eq. (1) to F˜K−N(√sρ),
(1)
and √s according to Eq. (12) depends on Re V , a self-consistent (SC)
ρ K−
procedure was applied, with 5–6 iterations proving more than adequate for
(1)
convergence. In this way a simple algorithm for constructing V from any
K−
7
given model for ˜K−N(√sρ) was formulated. As mentioned in the Intro-
F
(1)
duction, the SC V potentials constructed thereby were characterized by
K−
ignoring multi-nucleon absorption processes and by an imaginary part that
goes rapidly to zero towards the πΣ threshold. There were no free parame-
ters at this phase of the calculation and there was no coupling between the
real and imaginary parts of the potential beyond that provided by the input
−
K N amplitudes. In carrying out global fits to strong-interaction shifts and
(2)
widths data across the periodic table, a phenomenological potential V was
K−
added to the SC V(1), with parameters determined in a χ2 fit search with-
K−
(1)
out perturbing the prior SC determination of V . However, no compelling
K−
(2)
reason was given why V was excluded from the SC procedure.
K−
(2)
In the present work, a phenomenological term V is included from the
K−
(1)
outset in the SC procedure. Eq. (12) is thus modified, replacing V by
K−
(1) (2)
VK− = VK− +VK−:
(√sρ)atom Eth BNρ/ρ¯ 15.1(ρ/ρ0)2/3+ξK(ReVK−+Vc(ρ/ρ0)1/3) (MeV) ,
≈ − −
(15)
where the subscript atom indicates the limitation to kaonic atoms, thereby
alsosettingB =0. Thefirsteffectofthismodificationistointroduceimplicit
K
coupling between V(1) and V(2), since varying parameters of V(2) affects
K− K− K−
the √sρ energy argument of the in-medium ˜K−N(√sρ) and that of the
F
˜
underlying free-space fK−N(√sρ) amplitudes. Early tests of this approach
using the NLO30 amplitudes revealed [5] that this coupling is non-negligible
when theimaginary part ofthe single-nucleon amplitude drops sharply below
threshold, which was particularly the case with the ‘SE’ in-medium version
of model NLO30. The increased flexibility due to this coupling leads in the
present work, building on the IHW free-space amplitudes, to good fits to the
data, well beyond what was achieved in our earlier works [3, 4, 5].
Fig. 2 shows the functional dependence generated by the SC relation
−
Eq. (15) between subthreshold K N energies and nuclear densities for Ni
and Pb, calculated using the IHW-based global fit to kaonic atoms detailed
(1)
in Section 3. Compared with the earlier version [5] where only V appeared
K−
in the SC expression, Eq. (12), lower energies are now probed at the higher
densities and higher energies are probed at lower densities. Consequently,
for the more relevant region of 50% of the central nuclear density, the energy
downward shift remains unchanged at 40 MeV below threshold.
8
−10 E vs. ρ
V) −30
e
M
(
h Ni
Et −50
−
E
Pb
−70
−90
0 0.2 0.4 0.6 0.8 1
ρ/ρ
0
− −
Figure 2: Subthreshold K N energies probed by the K nuclear potential at threshold
as a function of nuclear density in Ni and Pb, calculated self-consistently according to
Eq. (15) from the IHW-based global fit to kaonic atoms specified in Section 3, see text.
2.4. Features of density dependence
Having introduced by Eq. (15) a relationship between the subthreshold
−
energy argument of K N amplitudes and their implied density dependence,
it is instructive to demonstrate the density dependence generated by the
WRW renormalization Eq. (9). Shown in Fig. 3 is the ratio of the in-
R
medium effective amplitude eff , Eq. (11), to the free effective amplitude
FK−N
feff , calculated as a function of nuclear density in Ni for the IHW ampli-
K−N
(2)
tudes in the absence of the V potential term. Note the logarithmic density
K−
scale used to highlight the slow convergence of to its low-density limit. For
R
example, Re =1.14 and Im =0.07 near 0.1% of the Ni central density ρ ,
0
R R
still far from their limiting values Re =1 and Im =0, respectively. This
R R
slow convergence is caused by the predominance of the Λ(1405) resonance for
densities roughly below 6% of ρ where Re exhibits hump structure with
0
R ˜
values exceeding 1, owing to the large negative values assumed by Re f
I=0
near threshold; the position of the Λ(1405) is marked here by the vanishing
of Re f˜ at √s 1415 MeV. At densities above 6% of ρ , Re 1, decreas-
I=0 0
≈ R≤˜
ing monotonically with density owing to the rapid increase of Re f below
I=0
˜
1415 MeV and leveling off when Re f has reached its (positive) maximum
I=0
value.
9
1.5
Re
1.0
0.5
eff
eff/ f
F
0.0
Im
−0.5
−1.0
10−5 10−4 10−3 10−2 10−1 100
ρ/ρ
0
Figure 3: Ratio of the in-medium effective IHW amplitude Eq. (11) to the free effective
IHW amplitude as a function of nuclear density in Ni.
A similar analysis shows that the Λ(1405) is directly responsible for Im
R
becoming slightly positive below a fraction of 1% of ρ and slowly converging
0
to its low-density limiting value of 0. However, it is not clear whether the
isospin-based WRW method of introducing medium effects into the free-
space IHW amplitudes, as applied here, is valid at such low densities where
−
the K Coulomb potential V becomes comparable with and even exceeds
c
V(1). In order to check sensitivities, χ2 fits were made also by imposing
K−
a transition to the free-space amplitudes at densities of 0.5, 1.0 and 2.0%
of ρ0, while retaining VK− down to densities 10−5ρ0 as normally done in
kaonic atom calculations. Below 10−5ρ , onl≈y V is retained, along with
0 c
changing the radial integration scheme from nuclear to atomic dimensions.
Weak sensitivities were found to the way the amplitudes were handled at
very low densities. Most of the results in this paper are for transition to free
amplitudes at 2% relative density.
3. Results and discussion
Here we report and discuss results of fits to kaonic atom data, using the
methodology outlined in the previous section. A phenomenological potential
(2) −
V representing K multi-nucleon processes is defined in Section 3.1, the
K−
10